Data-Driven Solutions to Protect the Environment

Blog /

Daniel Ho (left), Dan Palmer (right).
Daniel Ho (left), Dan Palmer (right).

The Regulation, Evaluation, and Governance Lab (RegLab) is working with federal and state environmental protection agencies to build a program for clean water and environmental compliance. The RegLab is one of six teams to receive start-up funding from Stanford Impact Labs for 2021-2023. 

RegLab’s faculty director Daniel Ho and Environmental Protection Agency (EPA) Attorney Dan Palmer describe the team’s work to use EPA data to increase compliance with environmental regulations designed to keep water clean and safe for everyone.  

What social problem are you working on?

DANIEL HO (Stanford): Environmental laws protect the air we breathe, water we drink, and land we inhabit. Federal, state, and local officials who are responsible for environmental governance face two urgent and complex social problems: rampant noncompliance coupled with insufficient capacity to enforce regulations. 

Noncompliance is pervasive. A third of hazardous waste generators violated hazardous waste regulations. Eighteen of the 25 largest coal-fired power companies were in serious violation of the Clean Air Act. Runoff from agriculture and improperly treated wastewater generate hypoxic conditions that suffocate marine life. Tens of millions of Americans are exposed to pollution hotspots and unsafe drinking water, disproportionately affecting poor, rural, and minority communities. 

At the same time, government resources and capacity for enforcement have dwindled, with inspections down nearly 50 percent over a 20-year period. Although the Environmental Protection Agency (EPA) houses one of the largest troves of data in government, federal and state environmental protection agencies lack the full capacity to use the data and state-of-the-art scientific methods to increase enforcement. 

Our team is partnering with federal and state regulators, environmental interest groups, and other stakeholders to address the social problem of environmental noncompliance and to pave the way for evidence-based environmental governance.

DAN PALMER (EPA): The Environmental Protection Agency (EPA) is the primary agency responsible for federal environmental policy and for coordinating enforcement and compliance activities. The agency focuses enforcement and compliance assurance resources on the most serious environmental violations by developing and implementing national program priorities, called National Compliance Initiatives. One of these initiatives aims to improve surface water quality and mitigate potential impacts on drinking water by reducing the national significant noncompliance baseline rate of 20.3 percent by half by the end of FY 2022. This goal, to cut the significant noncompliance baseline rate in half, is a “moonshot” goal. To achieve it, the EPA and the states will have to significantly change how they implement the Clean Water Act program. They will need to deploy new tools that are more effective in changing compliance behavior. In collaboration with Stanford RegLab, we are using scientific methods and EPA data to test and measure how much different compliance assurance tools change behavior and can help cut significant noncompliance with the Clean Water Act in half. 

What will the start-up lab funding and team approach help you do? 

DANIEL HO (Stanford): Our research is problem-driven and oriented toward partnership.  We believe university partnerships with the public sector have a critical role to play in revitalizing core government functions. Progress is most likely when the resources of the academic sphere work in tandem with government and public interest partners, who have deep domain knowledge and the opportunity to implement policy interventions. Our partnership with federal and state environmental protection agencies has enabled us to explore the most pressing and tractable problems in environmental compliance, and to determine how scientific research spanning disciplines -- including the social sciences, data science, computer science, geophysical science, and environmental science -- might address those needs.  With support from the Stanford Impact Labs, our team seeks to demonstrate how research-practice partnerships can drive science and technology forward to solve fundamental problems of public governance.

DAN PALMER (EPA): Our partnership with Stanford has enabled us to draw on machine learning, advanced research, and evidence-based practices to support the Environmental Protection Agency’s National Compliance Initiative to reduce the rate of significant noncompliance with Clean Water Act permit requirements. More specifically, this partnership is enabling us to roll-out national studies that examine the value of new approaches for improving compliance behavior.

What are you most excited about with this work?

DANIEL HO (Stanford): Michael Lewis’s “Fifth Risk” describes the risk of “the innovation that never occurs and the knowledge that is never created, because you have ceased to lay the groundwork for it.” We are excited to work to lay that groundwork.  

Developing a blueprint on how to strengthen government capacity for data-driven and evidence-based compliance is critical now because of the ways some agencies have been hollowed out. This work is necessary across federal and state government, whether to help prioritize scarce resources, to augment compliance capacity, to provide much-needed evidence about what works, or to invest in institutional readiness. We are excited to contribute to these urgent questions and a modern form of environmental governance. 

DAN PALMER (EPA): The EPA has set ambitious goals to improve environmental compliance. We are excited to have Stanford researchers’ support to drive  forward our strategic priorities and help bridge the capacity gaps of our agency. This effort is one of the first attempts by the EPA to use machine learning to meet its strategic goals and this collaboration has the potential to make a fundamental change in the way we and states approach environmental enforcement and governance. We look forward to our long-term partnership with this team and our continued work to develop data-driven solutions for environmental noncompliance problems.

Read more about the team members and impact lab’s work.